The recruitment of patients to trials in head and neck cancer: a qualitative study of the EaStER trial of treatments for early laryngeal cancer.

Research paper by D W DW Hamilton, I I de Salis, J L JL Donovan, M M Birchall

Indexed on: 22 Jan '13Published on: 22 Jan '13Published in: European Archives of Oto-Rhino-Laryngology


We aimed to investigate the factors contributing to poor recruitment to the EaStER trial "Early Stage glottic cancer: Endoscopic excision or Radiotherapy" feasibility study. We performed a prospective qualitative assessment of the EaStER trial at three centres to investigate barriers to recruitment and implement changes. Methods used included semi-structured interviews, focus groups and audio-recordings of recruitment encounters. First, surgeons and recruiters did not all accept the primary outcome as the rationale for the trial. Surgeons did not always adhere to the trial eligibility criteria leading to variations between centres in the numbers of "eligible" patients. Second, as both treatments were considered equally successful, recruiters and patients focused on the pragmatics of the different trial arms, favouring surgery over radiotherapy. The lack of equipoise was reflected in the way recruiters presented trial information. Third, patient views, beliefs and preferences were not fully elicited or addressed by recruiters. Fourth, in some centres, logistical issues made trial participation difficult. This qualitative research identified several major issues that explained recruitment difficulties. While there was insufficient time to address these in the EaStER trial, several factors would need to be addressed to launch further RCTs in head and neck cancer. These include the need for clear ongoing agreement among recruiting clinicians regarding details in the study protocol; an understanding of the logistical issues hindering recruitment at individual centres; and training recruiters to enable them to explain the need for randomisation and the rationale for the RCT to patients.

More like this: